Simple Classification Model for the Sonar Dataset

Template Credit: Adapted from templates made available by Dr. Jason Brownlee of Machine Learning Mastery.

For more information on this case study project, please consult Dr. Brownlee’s blog post at https://machinelearningmastery.com/standard-machine-learning-datasets/.

Dataset Used: Connectionist Bench (Sonar, Mines vs. Rocks) Data Set

ML Model: Classification, numeric inputs

Dataset Reference: https://archive.ics.uci.edu/ml/datasets/Connectionist+Bench+%28Sonar%2C+Mines+vs.+Rocks%29

The Sonar Dataset involves the prediction of whether or not an object is a mine or a rock given the strength of sonar returns at different angles. It is a binary (2-class) classification problem.

The baseline performance of predicting the most prevalent class is a classification accuracy of approximately 53%. Top results achieve a classification accuracy of approximately 88%.

The purpose of this project is to analyze a dataset using various machine learning algorithms and to document the steps using a template. The project aims to touch on the following areas:

  • Document a regression predictive modeling problem end-to-end.
  • Explore data transformation options for improving model performance
  • Explore algorithm tuning techniques for improving model performance
  • Explore using and tuning ensemble methods for improving model performance

The HTML formatted report can be found here on GitHub.

扭轉阿林斯基的規則

(從我的一個喜歡與尊敬的作家,賽斯 高汀

在“激進分子規則”的書中,索爾·阿林斯基提出了十三條原則,可用在於政治環境之內的總和為零的遊戲,以去打擊和擊敗政治敵人。

這種方法在任何問題上都經常被雙方利用,並且把文明的交流批的體無完膚。如果你很確定你是對的,而且你願意把你的對手的理念焚毀,那麼每個人遲早都會站在一座燃燒的環境裡。

如果我們能扭轉這些規則,那又會有什麼樣的後果呢?

  1. 讓員工去作事生產,它比光有金錢流動還更有效益。
  2. 挑戰你的員工去探索,學習,和去適應不確定性。
  3. 找到方法去幫助其他人找到更堅實的基礎。
  4. 幫助他人去編寫他們可用實現目標的規則。
  5. 按照你想要被對待的方式去對待別人。
  6. 不要光去批評。當教育能有幫助的時後,即使它不具有娛樂性質,也要這樣做。
  7. 堅持利用你的策略,雖然在別人對他們感到厭煩之後,只有當他們停止工作時才停止始用。
  8. 不時讓壓力停止是適當的。當人們無法一致地忽略它時,人們會關注你和你所尋求的變化。
  9. 不必要去威脅他人。你自己可以做或不做。
  10. 建立一支有能力和耐心的團隊來完成需要做的工作。
  11. 如果你一次又一次地把你的積極想法擺在首位,你會為其他人提高標準。
  12. 在花費大量時間為別人尋找問題之前,先解決自己的問題。
  13. 慶祝你的員工的存在,讓他們無拘束的去做更多的事情,讓他們成為隊友並邀請更多夥伴。只與機構持異議,讓他人有自己的意見。

The Three Curves

In the book, The Dip: A Little Book That Teaches You When to Quit (and When to Stick), Seth Godin teaches us why we should always strive to be “the best in the world.” Here are my reflections on the topics discussed in the book.

Being number one in your field matters and has tremendous benefits.

Think of the short head vs. the long tail.

Being on the top means people seek you out by default. We rarely have the time or opportunity to experiment or to try things out.

Being on the top also means leveraging scarcity to your advantage. Market rewards scarcity if what you offer is valuable.

But to get to the top, it will require you to get through the Dip (curve #1). Almost everything in life worth doing has a dip.

The dip separates the “aspiring beginner” from the “committed expert.”

The Cul-de-Sac (curve #2) can look pretty in a real-estate advertisement, but it is still a dead-end.

When you are progressing on a curve of Cul-de-Sac, nothing gets better or worse. It remains a status quo.

Sometimes we may even confront a Cliff (curve #3). It signals a nasty drop-off at the end of the curve.

If you find yourself in a dip on your way to something much more enlightening, keep at it.

Think of the dip as the belly of “The Beast” or Resistance in Steven Pressfield’s book “Do the Work.”

If you find yourself in a Cul-de-Sac or, worth yet, a curve of Cliff, you must quit and quit right now.

Continuing the curves of Cul-de-Sac or Cliff are simply a waste of time and resources.

Recognizing the three curves and quit when you must.

The End

In the book, Do the Work, Steven Pressfield talks about the self-sabotage we do to ourselves and what to do to overcome that self-sabotage. Here are my reflections on the topic in the book.

All important or worthy effort can boil down to three acts: the beginning, the middle, and the end.

After we slog through the tough beginning and the hopeless middle, we are nearing the end.

Develop the “Killer Instinct”

The beginning of the end is to ship.

Shipping requires a “killer instinct” because the act of shipping is to finally put the beast to rest, for this particular project.

Shipping takes courage and is not for the faint of heart.

“Fear of Success”

Marianne Williamson said the best. Also in “Akeelah and the Bee” and “Coach Carter.”

When we ship, we are exposed.

“Exposure”

When we ship, we set ourselves up for others to judge our work.

Exposure is like the empty space beneath a mountain climber – there is nothing there to break the fall.

If we do not ship when we should, the Resistance has beaten us.

We need to remind ourselves this:

“A Professional does not take failure (or success) personally.”

“Start (Again) Before You’re Ready”

The Resistance maybe temporarily defeated or subdued with this project, but it will be back on the next project.

The only way to keep beating the Resistance is to do what Pressfield has said all along.

“Stay Primitive,” “Swing for the Seat,” and “Trust the Soup.”

Simple Regression Ensemble Model for Boston Housing with Python

Credit: Template and study cases were adapted from blog posts made available by Dr. Jason Brownlee of Machine Learning Mastery.

For more information on this case study project, please consult Dr. Brownlee’s blog post at https://machinelearningmastery.com/regression-machine-learning-tutorial-weka/.

Dataset Used: Housing Values in Suburbs of Boston

ML Model: Regression, numeric inputs

Dataset Reference: https://archive.ics.uci.edu/ml/datasets/Housing

The purpose of this project is to analyze a dataset using various machine learning algorithms and to document the steps using a template. The project aims to touch on the following areas:

  1. Document a regression predictive modeling problem end-to-end.
  2. Explore feature selection options for improving model performance
  3. Explore algorithm tuning techniques for improving model performance

For this “Take-2” version of the project, we added the ensemble models to the exploration.

  1. Explore using and tuning ensemble methods for improving model performance

The HTML formatted report can be found here on GitHub.

你會要還是你將要

(從我的一個喜歡與尊敬的作家,賽斯 高汀

“你會要嗎?”幾乎總是不能喚起有用的信息。 那是因為大家都想做好人,並且想要保護著你的感受。“當然,如果你建立起了甲,乙,丙或,丁,我當然會考慮去購買它。”

但從另一方面,如果你問“你將要嗎?”,你馬上就會得到真相。“是的,今天我將要從你那裡購買它”

你可以做完世界上所有的研究,但是直到你有膽量去開口問直接的問題,你很難能確定任何事。

The Crash

In the book, Do the Work, Steven Pressfield talks about the self-sabotage we do to ourselves and what to do to overcome that self-sabotage. Here are my reflections on the topic in the book.

It is not enough to just get over the wall, everything inevitably “crashes” at some point.

“Ringing the Bell”

In SEAL training, you can ring a bell when you are ready to quit.

We all have the same bell, but will we ring it?

“Crashes Are Good”

We crash when we fail, but failure can be a good teacher.

It is up to us to learn from the crash and figure out the difference between what worked and what did not.

“Panic Is Good”

When we are a good streak or when the successful end is in sight, we sometimes panic.

Panic can set in when we are about to experience a change.

Our natural instinct kicks in and alerts us something big is about to happen.

When we panic, it is a sign that we are changing and growing.

“The Problem”

When problems arise, the professionals work the problem.

“A professional does not take success or failure personally,” said Pressfield.

The professional leverages the internal forces from the beginning of the project to power through, Stupidity, Stubbornness, and Blind Faith.

For a writer, working the problem may involve a re-write.

At last, we all solve the problem and get through the hell of Resistance.

Simple Regression Baseline Model for Boston Housing Price with Python

Credit: Template and study cases were adapted from blog posts made available by Dr. Jason Brownlee of Machine Learning Mastery.

For more information on this case study project, please consult Dr. Brownlee’s blog post at https://machinelearningmastery.com/regression-machine-learning-tutorial-weka/.

Dataset Used: Housing Values in Suburbs of Boston

ML Model: Regression, numeric inputs

Dataset Reference: https://archive.ics.uci.edu/ml/datasets/Housing

The purpose of this project is to analyze a dataset using various machine learning algorithms and to document the steps using a template. The project aims to touch on the following areas:

Document a regression predictive modeling problem end-to-end.

Explore feature selection options for improving model performance

Explore algorithm tuning techniques for improving model performance

The HTML formatted report can be found here on GitHub.